跳到主要内容
Learn Motion
探索课程产品能力价格

下载 Learn Motion

用手机扫描二维码,下载 App。

App Store

扫码从 App Store 下载 Learn Motion打开商店页面

Google Play

扫码从 Google Play 下载 Learn Motion打开商店页面
登录
Learn Motion

Why AI-Designed Drugs Haven't Changed Medicine Yet

1The Promise and the Puzzle2From Molecule to Medicine: The Journey a Drug Must Survive3Where AI Actually Helps in the Pipeline4The Prediction Gap: When a Good Molecule Meets a Real Body5The Long, Expensive Road of Clinical Trials6Money, Incentives, and the Business of Drug Development7Regulation, Evidence, and Trust8What Would Have to Change
The Prediction Gap: When a Good Molecule Meets a Real Body

A clean model environment next to a messy living body

2 / 4
Look at the left side first. That is the world the model lives in: a single target, a single molecule, and conditions it was trained on. Now look at the right side. That is what happens once the molecule enters a body. It has to be absorbed, carried through the bloodstream, survive the liver, reach the right tissue, and act on the target without disturbing the many similar processes around it. Each of those layers is a chance for the molecule to fail, and none of them appear in the model. That is the gap. The model is not wrong about what it measured — it just measured a much smaller world than the one the drug has to survive.
0:00 / 0:00

The model works in a controlled space: one target, one molecule, known conditions, and a dataset that defines what "good" looks like. A living body is not that space. A drug taken into the body must survive the stomach, cross into the bloodstream, avoid being broken down by the liver, reach the right tissue, and act on the target without disrupting the many other processes that resemble it. The body also contains thousands of other molecules, enzymes, and cell types that can interact with the drug in ways the model never represented.

This is why the same molecule can score well and behave badly. The model measured it against a target in isolation. The body measures it against everything at once. The diagram places the two environments side by side so the missing layers are visible: the model's clean target-and-molecule view on one side, and the body's layered reality — absorption, distribution, metabolism, excretion, immune response, and off-target interactions — on the other.

上一页2 / 4下一页

Learn Motion

生成课程,认真学会。

运营方:Wuhan Daoyin Technology Co., Ltd.

联系我们: [email protected]
隐私政策用户服务协议

© 2026 Learn Motion